Gibberellin Biosynthetic Inhibitors Make Human Malaria Parasite Plasmodium falciparum Cells Swell and Rupture to Death

Malaria remains as one of the most devastating infectious disease, and continues to exact an enormous toll in medical cost and days of labor lost especially in the tropics. Effective malaria control and eventual eradication remain a huge challenge, with efficacious antimalarials as important intervention/management tool. Clearly new alternative drugs that are more affordable and with fewer side effects are desirable. After preliminary in vitro assays with plant growth regulators and inhibitors, here, we focus on biosynthetic inhibitors of gibberellin, a plant hormone with many important roles in plant growth, and show their inhibitory effect on the growth of both apicomplexa, Plasmodium falciparum and Toxoplasma gondii. Treatment of P. falciparum cultures with the gibberellin biosynthetic inhibitors resulted in marked morphological changes that can be reversed to a certain degree under hyperosmotic environment. These unique observations suggest that changes in the parasite membrane permeability may explain the pleiotropic effects observed within the intracellular parasites.


Introduction
Malaria, caused by the genus Plasmodium, is one of the most common infections in the world responsible for about 1 million deaths per year. Artemisinin-based combination therapies (ACTs) are currently the first-line and only the best weapon against malaria [1]. Drug resistance, which rendered chloroquine and sulfadoxine-pyrimethamine infective for malaria control from 1970 to 1990s, remains as one of the greatest challenges facing malaria control today. Besides, ACT is 4 to 226 more expensive than common drugs and access is generally poor in African countries [2]. Also, although its clinical efficacy has not yet been compromised, there are recent reports that show the first evidence of artemisinin resistance [3]. Admittedly, the development of new inexpensive, effective and safe drugs with new mechanisms is strongly needed.
Plant-specific organelles and mechanisms in the phylum Apicomplexa, in which Plasmodium and other medically and veterinarily important pathogens are included, have been brought to focus as potential targets for new drugs since associated enzymes were found in plants and bacteria but not in animal metabolic pathways. Examples of these are plant-like vacuoles in parasite cells and the mevalonate-independent biosynthesis of isoprenoid in apicoplasts [4], [5]. The rationale was further strengthened with the demonstration that the apicoplast is essential for malaria parasite survival [6] and that metabolic pathways in the apicoplast are essential for parasite growth [7]. In addition, identification of inhibitors in these pathways might also result in synergistic drug combinations, which could have increased therapeutic value. The plant hormone abscisic acid (ABA) and ABA biosynthetic inhibitors have, likewise, been shown to affect parasite egress from infected host cells in Toxoplasma gondii [8].
In this study, we have preliminarily explored a wide variety of plant growth regulators, including plant hormones and corresponding inhibitors, to investigate their effect(s) on the growth of the most virulent human malaria parasite, Plasmodium falciparum. We have also used the apicomplexan parasite, T. gondii for comparison. T. gondii infects a broad spectrum of hosts and efficient drugs with low side effects and usable for human therapies are also highly needed. Plant growth inhibitors are commonly used in agriculture for years and have been synthesized in bulk, efficiently and cheaply, either naturally or artificially. Well-established manufacturing methods and facilities, as well as their safety profile (toxicity and teratogenicity) in animals, crops and humans are also available. Thus, plant growth inhibitors showing anti-apicomplexan activities might give valuable clues for prophylactic or therapeutic reagents effective for infectious diseases caused by protozoan parasites.
Prior to use and dilution at various concentrations, AMO-1618, chlorocholine chloride, prohexadione and FC-907 were dissolved in distilled water. INA and ancymidol were dissolved in DMSO and diluted in 0.1% DMSO. Paclobutrazol and uniconazole P were dissolved in ethanol and diluted in the culture medium.
Enantiomeric resolution of racemic INA (Wako) was carried out by high performance liquid chromatography (HPLC; LC-908, Japan Analytical Industry Co., Ltd, Japan). Elution order of enantiomers was determined by an ultraviolet absorption detector (SPD-10A, Shimadzu, Japan) using Chiralcel OD column [2560.46 cm, Daicel Chemical Industries, Japan, mobile phase: n-hexane-2-propanol (7:3, v/v), flow rate: 1.0 ml/min]. All resolutions were carried out at 26uC; detection at 275 nm. The assignment of the peaks was achieved by comparison with the synthesized (S)-form INA sample as previously described [9]: peak with shorter retention time correspond to the (R)-form and the peak with longer retention time to the (S)-form. Percentage purity of the products was calculated to be 97.0% and 99.5% for (S)-and (R)-forms, respectively, based on peak areas.
In vitro antimalarial assay of plant growth regulators Asynchronous P. falciparum 3D7 was used. Various concentrations of compounds in appropriate solvents (water, ethanol or DMSO) were prepared and added to 12-well plates. Starting parasitemia was at 0.1% in 2.5 ml culture medium. Growth was assessed after 72 h by percentage parasitemia using thin blood smears. The number of parasitized erythrocytes over a total of 3,000 erythrocytes was examined. Drug-free control cultures were run simultaneously.
For T. gondii studies, confluent Vero cell cultures were incubated for 2 days and infected with 2.5610 5 tachyzoites in RPMI 1640 medium containing 3% FCS using a 96-well plate. Tachyzoites were harvested after 2 days and b-gal activity was analyzed using a colorimetric assay [12].

Morphological effects of gibberellin biosynthetic inhibitors on P. falciparum
Tightly synchronized parasites within 4 h life span were prepared using 5% sorbitol treatment and percoll centrifugation. Synchronized parasites were treated with either 50 mM INA or 250 mM AMO-1618 from 0 h (ring), 20 h (immature trophozoite), 28 h (mature trophozoite) or 36 h (schizont). Giemsa-stained thinblood smears were prepared after 4, 8 and 12 h treatment. Digital imaging was performed on a HC-300 (Fujifilm, Japan) and representative parasite images are shown.

Fluorescence Microscopy
Thin-blood smears of infected erythrocytes treated with INA were stained with acridine orange (100 mg/ml). Fluorescence microscopy and confocal imaging were carried out using the Axioplan 2 microscope (Zeiss, German) and SPOT PS-BW CCD camera (Seki Technotron Corp., Japan). Filter sets for green fluorescence (green: nucleoli; emission LP515, excitation BP 450-490) and red fluorescence (red: cytoplasm; emission LP590, excitation 546/12) were used. Nile Red staining was carried out by addition of 1 mg/ml dye to the culture medium 1 h prior to microscopic analysis. Nile Red was excited at 546 nm and emission detected above 590 nm. To stain with rhodamine 123 and LysoTrackerH Red DND-99, 10 ng/ml rhodamine or 75 nM LysoTracker were added to the culture and incubated for 1 h. Confocal images were obtained with excitation above 546 nm and emission above 590 nm.

Electron microscopy
Intraerythrocytic parasites treated with 50 mM INA for 6 h, 250 mM AMO-1618 for 8 h or 0.1% DMSO for 6 h were centrifuged at 35006 g for 1 min and fixed in 2% glutaraldehyde in 16 PBS for 1 h. After washing in 16 PBS, fixed cells were treated as described [13]. Sections were viewed in a transmission electron microscope JEM-1011 (JEOL Ltd., Japan).
Effects of osmotic pressure on INA-treated P. falciparum P. falciparum in RPMI 1640 containing 10% human serum was grown to 5-10% parasitemia and synchronized by sorbitol treatment. After 21 h, 0.9 ml of early trophozoite-stage parasites was mixed with 0.1 ml of 16, 26, 36 and 46 PBS and treated simultaneously with 50 mM INA or 0.1% DMSO for 8 h (INA stock solution was prepared as 50 mM in DMSO and diluted to 1/ 1000 for assay). Parasitemia was determined using Giemsa-stained thin smears by a laboratory technician blinded to treatment assignments. Each treatment was replicated thrice and experiments were performed twice.
Volumes of 0, 0.1, 0.2 and 0.4 ml sterilized water was adjusted to 0.4 ml by mixing with appropriate amounts of RPMI 1640 medium containing 10% human serum for the different dilution series. Hyposmotic media were mixed with 0.6 ml of P. falciparum culture synchronized to early-stage trophozoites.

GC-MS analysis of sterols in P. falciparum
Parasites were cultured in 5 ml RPMI 1640 supplemented with 1% serum free AlbuMAXH II (Invitrogen). Parasitized erythrocytes were collected by centrifugation and hemolyzed in 0.54 ml of 0.15% saponin in RPMI 1640. Erythrocyte-free parasites were washed twice with 4 ml of 16PBS and suspended in 0.4 ml of 16 PBS. Geranylgeraniol at 50 mg was added to the suspension as an internal standard and extraction with 1 ml of chloroform was carried out 3 times by shaking the mixture at 37uC. Chloroform extracts were pooled and 1 ml solution was directly injected into the GC-MS (gas chromatography/mass spectroscopy) instrument (JEOL; JMS-Bu25 GCMate: ionization energy at 70 eV, filament current at 300 mA) with a DB-5 capillary column [0.25 mm i.d.615 m, film thickness of 0.25 mm (J&W Scientific, USA)]. The He carrier gas flow rate was 1 ml/min (constant flow), the injector temperature was 250uC, and the samples were introduced by splitless injection. After injection, the oven temperature program was held at 80uC for 1 min, then increased to 200uC at 30uC min 21 , followed by a further increase to 300uC at 15uC min 21 after which it was held constant at 300uC for 2 min.
For quantitative analysis, commercially available cholesterol and geranylgeraniol (Sigma) was detected by GC-MS and corresponding peaks were used as authentic standards for cholesterol and geranylgeraniol, respectively. Cholesterol was calculated based on ratio of peak areas with geranylgeraniol.

Antimalarial activity of gibberellin biosynthetic inhibitors
All inhibitors of gibberellin used in this study affect synthesis of gibberellin in plants ( Fig. 1). Generally, the inhibitors blocked P.
falciparum growth in vitro (Table 1, Fig. S1); although, results were admittedly different among the inhibitors as well as between Plasmodium and Toxoplasma. For example, inabenfide (INA) exerted strong activities in both P. falciparum and T. gondii, whereas prohexadione exhibited no cytotoxicity to P. falciparum even at 500 mM. Likewise, ED 50 s of some inhibitors varied greatly between P. falciparum and T. gondii, such as AMO-1618: 10.362.22 mM in P. falciparum and .1000 mM in T. gondii. Apparent difference in effective inhibitor concentrations between the two parasite species could be species-specific barrier on membranes for drug uptake or inhibitory mechanisms in P. falciparum and T. gondii [14].
Some of the tested inhibitors have an asymmetric carbon, and their enantiomers are known to differ significantly in their biological properties [15]. (S)-forms are biologically more active than (R)-forms in higher plants [16]. In order to examine the chirality effect of these inhibitors, we separated enantiomers from commercially available INA by HPLC using a chiral column (Fig.  S2) since marketed INA is racemic. The (S)-form was slightly more potent than the (R)-form, as demonstrated by ED 50 values of 2.7860.51 mM and 6.3260.81 mM, respectively in P. falciparum (Table 1). Notably, the (R)-form was more effective in inhibiting growth of T. gondii.

Morphological changes induced by gibberellin biosynthetic inhibitors
We also evaluated morphological and physiological effects induced by the gibberellin biosynthetic inhibitors. In what appears to be the center of the parasite, a Giemsa-unstained region was prominent within 4 h after treatment with INA ( Fig. 2 B). The ''haloed'' parasites swelled and ruptured within 12 h of culture in immature trophozoite-stage. Rings were seldom found when early trophozoite-stage parasites were treated with INA. When INA was applied at late stages of the parasite, i.e. in mature trophozoites and schizonts, a similar swelling was also observed although the stained periphery appears broader with merozoites bloated in appearance and few in number (Fig. 2 C, D). Ring-stage treated parasite cultures were usually appliqué forms (rings appearing on the periphery of the erythrocytes) (Fig. 2 A).
Another type of gibberellin biosynthetic inhibitors, AMO-1618 and FC-907, which prevent cyclization reaction in the early steps of gibberellin biosynthesis ( Fig. 1) [15], [17], also caused parasite cells to swell. Swelled parasites were found within 8 h and 24 h after AMO-1618 treatment (Fig. 2, S3), but the altered morphology on schizonts was different: parasites appear abnormally disordered inside erythrocytes (Fig. 2 C, D). A similar morphological change in Plasmodium cultures were observed when greater than 2 mM INA were used (data not shown). Notably, both INA and AMO-1618 did not appear to damage erythrocytes, because P. falciparum could infect erythrocytes that had been treated for 24 h with 50 mM INA or 250 mM AMO-1618 without any significant inhibitory effects (Fig. S4).
At 2 mM, INA-induced morphological changes were microscopically visible at 24 h in asynchronous parasite cultures. In order to determine if INA and AMO-1618 targets a specific stage during intraerythrocytic development, we increased concentrations of INA to 50 mM and AMO-1618 to 250 mM. Both inhibitors at these concentrations provoked morphological changes more rapidly than concentrations at their ED 50 s (Fig. S5). The use of increased concentrations of inhibitors also allowed us to measure the intrinsic sensitivity of the parasites in vitro, and provided clues to the maximal response that can be produced by these inhibitors, as well as their toxicity to the human erythrocytes. Even after 24 h incubation, use of 0.1% DMSO alone gave no effects on growth and morphology of malaria parasites. Parasite swelling and rupture was also observed on P. falciparum cultures treated with other gibberellin biosynthetic inhibitors, such as paclobutrazol and uniconazole P (Fig. S5). These inhibitors, as well as INA, are known to block cytochrome P450-dependent monooxygenases (CYP701A) in plants [15]. Interestingly, the time and the concentrations necessary to provoke morphological changes were different among these compounds, presumably due to differences in solubility and lipophilicity.

The 'haloed' trophozoites
The ''haloed'' parasites were stained with various fluorescent probes to visualize intracellular components. The center of the trophozoites was not stained with acridine orange (Fig. 3 A). Both nuclei and cytoplasm appear to be on the periphery, as seen in the Giemsa-stained images. Nile Red, used to visualize various lipidrich compartments such as membranes, endoplasmic reticulum (ER) and lipid bodies [18], revealed a simpler staining pattern surrounding what appears to be fewer merozoites in contrast to the control. Merozoites also appear bloated with intense spots of fluorescence (presumed as lipid bodies, Fig. 3 B). In normal mature stage trophozoites, rhodamine 123 showed thread-like branched mitochondria, characteristic of the mitochondrial development at the later stages where each daughter merozoite receives a branch of the parent organelle (Fig. 3 C) [19]. In INA-treated parasites, however, branched mitochondria were rarely seen. Although fluorescence intensity in the mitochondria started to decrease after 2 h, parasites still have significant fluorescence over the entire parasite cytosol, except in the food vacuole, similar to untreated parasite cells. This observation suggest a membrane potential generated by the parasite's outer plasma membrane [20], [21] and that INA effect likely induced a specific reduction in the mitochondrial membrane potential alone.
These differences in the staining patterns were not evident when ring-stage parasites were observed (data not shown), which suggest that various membranes were remarkably affected by INA only when parasites reached the trophozoite stage. The timing is coincident at a time when the parasite starts to actively generate membranes needed for growth and schizogony [22].
Electron microscopy also revealed fine intracellular differences at the trophozoite and schizont-stage parasite (Fig. 4 A, B). When trophozoites were treated with INA or AMO-1618 ( Fig. 4 A), membranes of the nuclear envelope and ER were thicker and intensely stained compared to control parasites. Conspicuously unfamiliar spaces appeared around the ER and daughter nuclei ( Fig. 4 A, INA and AMO-1618 panel). Using asynchronous cultures, these spaces were more readily observed at mature-stage parasites than early trophozoites (Fig. 4 B, b and f vs. c and d). No differences were evident in the plasma membranes of both host and parasite cells during the intraerythrocytic life cycle (Fig. 4 B). These results suggest that the gibberellin biosynthetic inhibitors appear to affect components of various intracellular membranes of the parasites, and the effects were especially evident during schizogony.

INA effect and osmotic pressure in parasite cells
Various stressful conditions were applied to INA treated parasite: use of hydraulic pressure, ultrasound, pH (5-9) and temperature changes (4-37uC); but no differences were observed when compared to control parasites (DMSO-treated). However, different responses were seen when trophozoite-stage parasites were under various osmotic stress. Hyperosmotic conditions were achieved by incubating with 10% (v/v) 16, 26, 36 or 46 PBS. For control, untreated parasite cultures, relative parasitemia drastically decreased with increasing hyperosmotic environment (Fig. 5 A). However, relative parasitemia in the INA-treated Between normal and reversed PBS in the different treatments, no significant differences were observed in both parasitemia and morphology (Fig. S6). Exogenously added Ca 2+ (100 mM) to cultures did not help reverse the 'haloed' effects of INA in trophozoites (Fig. S7). Hyposmotic stress enhanced INA's effect in treated parasites: early trophozoite-stage parasites swelled even more by the addition of 10% (v/v) water to the culture medium (Fig. 5 B). At 40% (v/v) water, trophozoites swelled to almost the same volume of the infected erythrocyte. Decrease in parasitemia was also statistically significant compared to the control. These observations suggest that INA treatment might affect membrane permeability; water or ion influx might cause swelling of the parasites.

Isoprenoid biosynthetic pathway in apicomplexan parasites
Parasite cultures were incubated with gibberellin A 3 and A 4 , the most effective gibberellins, with or without gibberellin biosynthetic inhibitors. Of note, gibberellins did not affect growth of parasites nor overcome the effects of the inhibitors (data not shown), suggesting that the inhibitors could not disrupt gibberellin biosynthesis in P. falciparum or malaria parasites do not utilize gibberellins.
Gibberellins are diterpene (isoprenoid), ubiquitous and essential compounds by themselves or as materials of secondary metabolites and signal molecules in animals, plant and bacteria [15]. Biosynthetic pathways of isoprenoids are very complicated and parallel pathways for the biosynthesis of the universal 5-carbon building block for all terpenoid compounds exist. A huge variety of sterols and terpenoids are synthesized via isoprenoid metabolic pathways in a species-specific manner, although all pathways utilize farnesyl pyrophosphate as a starting molecule generated in the mevalonate pathway and the 2-C-methyl-D-erythritol 4phosphate (MEP) pathway ( Fig. 1) [23]. Gibberellin biosynthetic inhibitors have been known to block other enzymes leading to the production of sterols and terpenoids. We then analyzed P. falciparum treated with INA and AMO-1618 by GC-MS. Interestingly, any sterols or terpenoids besides cholesterol derived from human erythrocytes were not detected, and there were no significant differences in the level of cholesterol between control and INA-or AMO-1618-treated samples (Fig. 6).

Discussion
Efforts on malaria control have been complicated by emergence of drug resistance. With the apicoplast considered to originate from an ancestor with an endosymbiotic alga, and that many nonphotosynthetic processes are shared peculiarly with plants, plant growth regulators are of particular interest for pharmaceutical leads on the assumption that the molecular target may not be present in humans. The present study clearly showed some inhibitors that antagonize gibberellin biosynthesis in plants are effective to both P. falciparum and T. gondii (Table 1). Gibberellin is one of 5 major plant hormones known to exert various effects on physiological events, particularly germination and seed bearing by promoting cell elongation [24]. 126 different gibberellins are known to occur in plants [15], and they are synthesized via the mevalonate pathway and the MEP pathway. Interestingly, gibberellin has been utilized to induce development of parthenocarpous fruits in some fruits such as grapes. On the other hand, gibberellin inhibitors have also been widely used for agriculture, e.g. for dwarf plants without harmful effects to either plants or animals.
All gibberellin inhibitors tested in this study block biosynthesis of gibberellin in plants: AMO-1618 and FC-907 block synthesis of entkaurene (an intermediate during the synthesis of gibberellin) by inhibition of copalyl-diphosphate synthase (CPPS) and ent-kaurene synthase (KS, Fig. 1) [15]. Next to AMO-1618 and FC-907 steps, INA, paclobutrazol, uniconazole P and ancymidol interrupt entkaurene oxidase (KO), a member of the cytochrome P450 monooxygenases (CYP) oxidating ent-kaurene into ent-kaurenoic acid. Azole antifungal such as fluconazole and diniconazole are analogous compounds to these inhibitors and interrupt synthesis of ergosterol [25]. They have an asymmetric carbon and their enantiomers are known to differ significantly in their biological properties [16], [26]. (R)-forms demonstrate stronger activity than (S)-forms in terms of fungicidal effects, whereas the latter is functionally active and the (R)forms showed only residual activity with regard to blockage in plant gibberellin biosynthesis. We tested the stereochemical selectivity of INA and found that the (S)-form inhibited growth of P. falciparum .2 times stronger than (R)-form but vice versa in T. gondii ( Table 1). The stereochemical discrimination was, thus, similar to plants than to fungi in P. falciparum, although the (S)-form was reportedly 1000 times more potent than (R)-form in growth retardation of pumpkin seedlings [26]. Differences may be due to discrepancies in protein structures that relate to membrane permeability or translocation of the enantiomers to the site of biosynthesis in such distant species.
The observed decrease in parasitemia in INA-treated cultures correlated with remarkable morphological changes: swelling and rupture of P. falciparum cells, particularly when drug treatment commenced during early trophozoite stage (Fig. 2). Trophozoitestage parasite cells appear to be 'haloed', the center of the cell remains unstained by Giemsa and acridine orange. Both nuclei and cytoplasm were seen on the periphery of the parasite cells. These data suggest that some substances that are neither acidic nor alkaline might accumulate in the unstained parts.
We also observed the intracellular structures in INA-treated cultures using different dyes and found that various intracellular organelles, especially organelles synthesized during schizogony (such as mitochondria, membrane network and food vacuoles) are morphologically abnormal or compromised after short treatment with INA (Fig. 3). Membrane potential in the mitochondria was reduced rapidly within 2 h of INA treatment. Organelle membranes such as those in the ER and nuclear envelope appeared thicker and gaps were observed around such organelles under electron microscopy (Fig. 4). From these observations, it appears that gibberellin biosynthetic inhibitors damage the parasite internal organelles, compromises parasite membranes and results to abnormalities in the treated parasites.
In order to further probe for clues that can shed light into the mechanisms for such changes in the membranes, we exposed INAtreated parasite cells to various stresses, and found that osmotic pressure modified the effects of INA. Treated parasite cells under hyperosmotic conditions became less swelled and viability was significantly restored (Fig. 5 A). In contrast, the addition of water in the culture medium enhanced expansion and rupture of parasites (Fig. 5 B). These results demonstrate that INA treatment might affect permeability of membranes, and consequently result to uncontrollable influx of water or ion into parasite cells. No abnormalities in the plasma membrane structure were observed by electron microscopy (Fig. 4), although this cannot be completely ruled out since parasites with severely compromised plasma membranes are very likely susceptible to rupture or bursting during various treatments for microscopy.
In order to maintain the various functions and stability of plasma membranes, different organisms synthesize several specific components for their biological membranes. For example, sterols, a vast family of isoprenoids, take pivotal role with mammalian and fungal cells generally synthesizing one major sterol, cholesterol and ergosterol, respectively [23], [27]. Plants, on the other hand, are known to produce a characteristically complex sterol mixture. As many as 61 sterols and isoprenoids have been identified in a single maize seedling [28]. Prokaryotes produce hopanoids, pentacyclic isoprenoids, by direct cyclization of squalene ( Fig. 1) [29]. The functions of the hopanoids were shown to be equivalent to sterols: functioning as membrane reinforcers affecting membrane permeability and fluidity. In protozoa, there has been one report that Tetrahymena synthesizes tetrahymanol, a quasi-hopanoid, together with small amounts of diplopterol [30], [31]. Both components were incorporated into its membranes.
Enzymes that are involved in gibberellin biosynthesis are also of importance in the formation of such sterols and other membrane constituents. Change in the level of these substances after treatment with inhibitors has been often observed as side activities for these inhibitors [15]. For example, AMO-1618 inhibits syntheses of phytosterols and cholesterol, by blocking cycloartenol synthase (CAS) in plants and lanosterol synthase (LS) in animals, respectively [32].
We analyzed P. falciparum treated with INA and AMO-1618 by GC-MS in order to identify sterols or isoprenoids in malaria parasites, but we could not detect any isoprenoids, besides cholesterol (Fig. 6). We might be able to hypothesize that responsible components could be too low or structurally unknown to detect by GC-MS in parasite cells. Data about syntheses of isoprenoids in protozoa, especially in Apicomplexa, remains nil. Intraerythrocytic parasites are known to intake large amounts of erythrocyte contents with surrounding membranes that are cholesterol rich, however, cholesterol is greatly depleted in parasite membranes (Fig. 6 A) [33]. How parasites transport and dispose of excess cholesterol within the cells remain a puzzle. Additionally, does lack of isoprenoids make membranes leaky even in intracellular parasites? A previous study has demonstrated by differential scanning calorimetry and electron spin resonance the association of gibberellin molecules with phospholipid membranes thereby altering the fluidity or viscosity of the lipid bilayer [34]. Clearly, further studies on gibberellin biosynthetic inhibitors could shed light if these compounds could perturb lipid bilayers or alter properties of the membrane phospholipid. The gibberellin biosynthetic inhibitors are known to be specific for isoprenoid biosynthesis in plants. We searched amino acid sequences and conserved domain sequences essential for these enzymes using PlasmoDB, but no sequences with significant homologies were found. This may imply the uniqueness of isoprenoid metabolic pathways in Apicomplexa, and consequently, provide potent targets to develop novel therapeutic agents.   the counter was blinded to sample identities. Error bars represent the standard deviation of three independent experiments made in duplicate. Data are normalized relative to those for the control treated with 16 PBS and 1 ml/ml DMSO. (TIF) Figure S7 Influence of Ca 2+ on the effects of INA in P. falciparum. Parasites were synchronized by 5% sorbitol treatment and treated with 1 ml/ml DMSO or 50 mM INA with or without adding 100 mM CaCl 2 after 18 h of incubation. The parasites were examined at 6 h of the treatment by thin blood smears and staining with Giemsa. Scale bar indicates 3 mm. Each panel shows a typical morphology of trophozoite-stage parasites in each treatment. (TIF)